Using latent class analysis to identify treatment‐use subgroups among parents of children with autism spectrum disorder

Abstract

Among parents of 2,582 children (ages 4-17 years old) with autism spectrum disorder (ASD), we used latent class analysis to identify subgroups and profiles of treatment users and included annual household income in the specification of the models, then described characteristics of each subgroup. Based on three indicators of fit (Akaike's Information Criterion, Bayesian Information Criterion, and Lo-Mendell-Rubin), six latent classes of treatment users emerged. Subgroups included users of: (a) mostly private and school speech and occupational therapies; (b) nearly all treatment types; (c) mostly speech and occupational therapies, plus intensive behavioral and "other" treatments, but little medication use; (d) private therapies almost exclusively; (e) primarily psychotropic medications; and (f) mostly school-based therapies. Income significantly predicted class differences for all but one latent class. Probabilities of families' lifetime use of nine treatment types varied depending on latent classification. Proportions of families reporting having observed children's developmental regression were largest in those with the highest overall treatment use, and these children also had the lowest cognitive and adaptivefunctioning scores and the highest ASD symptom scores. Understanding patterns of treatment use among families of children with ASD is an important first step in enhancing treatment-related selection and implementation. Autism Res 2019, 12: 843-854.Lay Summary: We identified six different groups of treatment users to help explain patterns in treatment implementation among parents of children and adolescents with autism. These included families who used: (a) mostly used private and school speech and occupational therapies; (b) nearly all treatment types (private and school therapies, intensive behavioral, biomedical, psychotropic medications, and other treatments); (c) mostly speech and occupational therapies, plus intensive behavioral and "other" treatments, but little medication use; (d) private therapies almost exclusively; (e) primarily psychotropic medications; and (f) mostly school-based therapies.